Actually - Herve reminded me of how I used to estimate when I was really new to the biz. I'd spend quite a bit of time trying to figure out how long it would take me. I'd look over the drawings and say, it shouldn't take too long to do this, or that. Then I would get my number and multiply it by 2. I found that it always took me about 1.5x as long as I thought it would when I first started out. The 2x rule worked out really well for me. Sometime I would end up really under estimating and loosing a small amount, and other times I made a bit extra, which made up for it.
Also I always ended up a quote, usually sent by e-mail, by saying "give me some feedback on the amount and I'm sure we can work something out"
That way if it seemed too high for them I could decide if I wanted to drop the price a bit, and maybe cut corners a little bit on the final product.
I always quote an upset limit. Give them a range of hours and set an absolute high limit. As an example I would say it would take between 32 - 40 hrs and say I wouldn't charge any more than 40 but if it took shorted amount of time, I would charge that. It gives the client an idea of a low price, but invariably there are always changes and the higher price is always used.
Cheers Peter.